Introducing
Your new presentation assistant.
Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.
Trending searches
“The key to our product is the algorithms, and they’re proprietary,” said Jeffrey Harmon, Northpointe’s general manager.
“We’ve created them, and we don’t release them because it’s certainly a core piece of our business.
It’s not about looking at the algorithms. It’s about looking at the outcomes.”
intellectual property opposition
How do we reconcile the legitimate intellectual property rights exercised over algorithms with the explainability issues that arise when they are deployed in the realm of governmental or public interest decision-making?
semantic sophistication
Deep learning, reinforcement, neural networks
multiple layers of representation
Pasquale, Diakopoulos
AI as a set of techniques across
different applications
Law is important, but other governance tools
have to be considered
re-coding of the rule of law
responses might vary across sectors and jurisdictions
What does society
have to do with all that?
Further Developments
We all know that a constitutional or EU Law test will not ban automation just because there is no man in the loop, but only over prejudice (even if potential) or concrete harm
No framework for impact assessment, as in Environmental or Competition
- intellectual property rights over algorithms and data sets with significant public relevance
GDPR, Art. 22
- Environmental Law And Competition Law use impact assessment tools that might help
A
PLPD Brasil, Art. 19
transparency
accountability
Why now?
fairness
complexity
explainability
What?
and not before?
automated law-enforcement
legal
usability
"the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use."
ISO 9241-11, Geneva, 1988
Moment of decision (high semantic effort)
x
Moment of enforcement (lower semantic effort)
potential applicability
Identify stages of technological support for law-enforcement procedures
(manual, instrumental, automation)
Teleonomic
X
Teleological
real-time law enforcement (man in the loop)
x
unmanned law-enforcement (man out of the loop)?
Moment of enforcement (lower semantic effort)
Teleonomic
instrumental or automated
Solutions are unstructured, not systematic, developed case by case and technology by technology
Maybe instances are not yet legally significant, but adoption and massification will make Law incorporate them
real-time or
unmanned
intense, fast, deep
and wide disruption
fully implemented
considerable direct impact
legally sensitive
fully implemented
average direct impact
not so legally sensitive
legal order
if we weren't just imagining
Imagine
but now
research project
massive direct impact
legally sensitive
context
fully implemented
massive direct impact
legally critical
under development
considerable direct impact
legally critical
algorithms
in use
sensors
unknown direct impact
legally critical
and
data
from running stock markets to the criteria of my VISA appointments, from my bank credit to the offers I get as a consumer, from my risk assessment by health and insurance companies to which neighborhoods to direct police resources, from being a suspect in a crime to public policy and interest decisions
accountability
compliance
Air Gesture
ICIL 2016
7th International Conference on Information Law and Ethics ΙCIL
Broadening the Horizons of Information Law and Ethics
Α Time for Inclusion
Pretoria, South Africa
February, 2016
automated law-enforcement
legal
usability
"the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use."
ISO 9241-11, Geneva, 1988
Moment of decision (high semantic effort)
x
Moment of enforcement (lower semantic effort)
potential applicability
Identify stages of technological support for law-enforcement procedures
(manual, instrumental, automation)
Teleonomic
X
Teleological
real-time law enforcement (man in the loop)
x
unmanned law-enforcement (man out of the loop)?
Moment of enforcement (lower semantic effort)
Teleonomic
instrumental or automated
Solutions are unstructured, not systematic, developed case by case and technology by technology
Maybe instances are not yet legally significant, but adoption and massification will make Law incorporate them
real-time or
unmanned
iGorithm
You are what you browse
Why Law?
Where?
Cláudio Lucena
it is by all means unacceptable to despise that the development and the incorporation of digital technologies in everyday human life both have paramount impact over fundamental and personality rights, as well as over other non-economical values duly protected by Law."
Cláudio.Lucena@uepb.edu.br
Descriptive
Concentration on structural informational elements (sensors, data, algorithms), rather on systems (devices or systems, drones, robots, etc.)
Review the context of the research from a Law and Technology approach, present concepts, examine classifications and interdisciplinary features, establish relationships and explore alternatives, using existing theoretical references to justify the research choices made from within available options
Empirical
Case study to identify characteristics of selected technologies which are currently in use to automate law-enforcement measures, which legal elements are considered and the policy-making process
Holistic view, general theory (as opposed to specific liability, military use, etc.)
Experiment?
regulatory law
consumer protection
property
data protection
contracts
Normative
Being possible to identify common features among studied instances, consider the viability of a legal dimension of the concept of usability in the realm of automated law-enforcement, as an element of a framework to assess the legality, maturity and adequacy of technological solutions deployed into legal coercive measures
intellectual property
intergenerational data protection framework
behavioral nudging of the networked self
deep learning normativity
reciprocal transparency
usability
delisting
automated law-enforcement
internet governance
world scale data operators
algorithmic accountability
international jurisdiction
non deterministic
but
nature
legally relevant to govern algorithms
No
challenges
measures
Fine-tuning
Jacky Alciné
Word Embedding
YouTube ContentID
How is it possible to design an accountability framework when deploying deep learning mechanisms? (auto-feedback, decision, even if not necessarily legal)
Jurisdictional compatibility and
interoperability of automated coercion
Private enforcement (delist, RTBF)
X
Public enforcement
(administrative instances X court jurisdiction)
Compulsory mass sensoring legal?
Shift from compliance to individual monitoring?
What interests can it benefit?
Under which conditions?
Which jurisdictions would adapt?
(comparative approach)
Which scopes of law would fit in?
(case approach, law and economics)
Collective Rights and Digital Content, Cláudio Lucena, Springer, 2015.
VOTRE TIRE ICI
WS #264 Automated Guardians of the Good? Algorithms impact in the exercise of rights
Catherine García, Cláudio Lucena, Martín Silva, Rachel Pollack, Sizwe Snail
Geneva, Switzerland
December 20th, 2017
par http://www.prezcreation.com
3D TEMPLATE
not to mention